Host Engineering Forum
General Category => ECOMs and ECOM100s => Topic started by: todd1313 on February 14, 2008, 03:55:43 PM
-
Every H2-ECOM100's that we have on site (3 of them) gives an error after some hours of running...maybe 4 to 12 hours. The H2-ECOM's that we have seem to be running fine with no problems. I have the subnet mask and gateway set according to the IT dept. recommendations. I can find no reason for it to trigger the error. Usually I have LabVIEW reading the data from it but it will give the error even with LabVIEW off. Are there any other events that could trigger the error?
Thanks for any help!
-
I assume that when you say "gives an error", you mean the red ERR light comes on on the H2-ECOM100s. More than likely, this is caused by some form of electrical noise (radiated or direct). Please see http://www.hosteng.com/FAQFiles/ECOM.htm#FAQ0045 and see if this applies to you.
-
I am having the same problem with a H2-ECOM100 card locking up w/ the red error light on. This one has caused considerable heartburn, as it is a remote site that requires a 20 minute drive to get to.
A short system overview. Our Ethernet network contains one DL205 Master, one DL06 slave, one DL205 slave (the problem child), with a C-More HMI that is accessing data from the DL205 Master and the DL06 Slave. These items are all connected via individual cables to a common switch. The control room data that we need is pulled from the master, via Modbus radio, so the Ethernet network is local only. At one point we were talking RX/WX broadcasts, the old way, but I am now using a peer to peer table in the master PLC. I do notice a slight hesitation once in a while on the problem ECOM card, but can't nail it down to any specific event. If I pull either the DL06 or the C-more out of the network, the hesitation goes away. I don't know if this would have anything do do with the fault?
After going to your FAQ0045, I checked off what we are doing against your recommendations.
(1) We use Cat 5e jumpers and cables.
(2) Thinking that electrical noise might be an issue, we took a temporary cable directly out of the PLC cabinet door to our switch to minimize interference. The problem stayed with us.
(3) I was wondering what noisy I/O might be considered to be? We have a 260 CPU in slot 0, a D2-08ND3 input card in slot1, an F2-08TRS output card in slot 2, an F2-04AD-1 analog input card in slot 3, a F2-02DA-1 analog output card in slot 4, and the H2-Ecom100 Ethernet card in slot 5. All of the I/0 is 24 VDC with interposing relays for higher voltage loads. The PLC base is 24VDC.
(4) The ethernet cable was originally routed with other 24VDC wiring, but the solution in #2 above would have taken care of that.
(5) We have a ground wire pulled from my main service to a PLC cabinet grounding point. The PLC is grounded to this point.
I am assuming that line filters are not needed on a 24VDC base as the 24VDC power supply would act as a filter of sorts.
Looking back over our system trends, I noticed that the card faults only when the chlorine generation process is running, but is at very random intervals- 1 hour to multiple days.
There is a large DC rectifier cabinet mounted directly below the PLC cabinet. Would it be possible that the non-linear loads would affect this card, even though the wiring is not in close proximity to each other?
I also thought that heat might be an issue. The room gets up to 90 deg F at times, but this is well within specs.
Last, but not least. According to a further FAQ, it looks like there is no way to reset the ECOM but to cycle power?
If anybody has any suggestions, I would appreciate the input.
-
Hi watertech2,
On the problem-child H2-ECOM100, can you temporarily pull the card and look on the label on the PWB (backside), and tell me the information contained there? Should look something like:
Host Auto. Products
H2-ECOM100 4A
0707 5VDC
00.E0.62.20.91.69 MFG:VIN
The reason I'm asking, is that I still believe it is a noise-related issue of sorts (i.e. hard to troubleshoot). I'm wanting to see what version of hardware it has. In the example above that is indicated by "4A".
-
This is interesting. I have an installation where a 260 is used as a Modbus TCP slave to an MMI, which experiences recurring comms problems that fix themselves almost instantly. The customer is a kind of eclectic guy with an unusual MMI running on an office grade PC with Ethernet cables the guy made himself, so I had sort of written off the problem as being due to one of those factors. Can you give us some more detail on the "problem child" status of the H2-Ecom100? In the meantime, I'll have the guy check the hardware revision level of his card.
-
OK, my customer's is:
Host Auto. Products
H2-ECOM100 3D
0307 5VDC 33ZL
00.E0.62.20.84.02 MFG:VIN
Do I have to shoot it? Him?
-
Controls Guy & watertech2,
The H2-ECOM100s Revs 1-3 (Controls Guy's is Rev3 indicated by "3D") have shown the same robust performance of other Host products in most applications in the field. In a couple of apps where high voltage switching (e.g. 220 vac) was very close in proximity we did find a weakness and so we made further improvements to the hardware and called it Rev4.
With the higher frequency of operation of the ECOM100 (versus ECOMs; 100Mbps vs 10Mbps) the challenge of letting in the high frequencies that need to be there and keeping out those that don't is a bit more difficult. Rev1-3 did a great job and there are 1000s of those successfully operating in the field.
However, not being content with a product that worked in most places we made a running change about a year ago to insure that it works in all places. :)
What you guys may be experiencing could be related to this issue. I'll PM/e-mail you and get more details, or you can PM/e-mail me.
-
I was on site yesterday, as the card had locked up again. I rerouted the temporary cable up and away from the rectifier cabinet as much as possible (the control and rectifier cabinet are only separated by a foot or so- possible radiated EMF?).
I didn't catch your post in time to get the card revision #, but will probably be out close to that site sometime early next week. Luckily, this part of the process is stand alone and so far we are just using the ethernet connection for monitoring only. Thanks for the input!
-
OK, here's the skinny on the hardware revisions.
The problem card in the 260 rack is:
H2-ECOM100 3D
0507 5VDC
The card in the 250 rack is:
H2-ECOM100 3C
0406 5VDC
My spare card is:
H2-ECOM100 3D
0507 5VDC
This clears the waters a little further. At one point, I had pulled the original problem card, replaced it w/ my spare, and saw the same behavior. I then put the original card back in as it didn't solve my problem. If the problem is in the hardware, it wouldn't have- they are the same revision.
The ECOM-100 card with the 3C hardware installed at the same site, different room, is rock-solid. This, along with a few other installations that we have, proves that these older cards will work for most of our applications. This is the first time that we have installed an Ethernet card on this process.
Greg, at this point, it looks like I might have to buy a new card. Do you have any other suggestions for me at this point?
-
This one being a prior rev hardware failure, we'll take care of it. Sorry for the pain...thanks for the patience.
Mr. Greg, please provide the appropriate remedy via pm/email.